expanded: Biological Overview | Regulation | Developmental Biology | Effects of Mutation | References

Gene name - expanded

Synonyms -

Cytological map position - 21C2--3

Function - signaling

Keywords - imaginal discs, proliferation, tissue polarity, tumor suppressor, Fat signaling pathway

Symbol - ex

FlyBase ID: FBgn0004583

Genetic map position - 2-0.1

Classification - Band 4.1 protein with three potential SH3-binding sites

Cellular location - cytoplasmic

NCBI links: Precomputed BLAST | Entrez Gene
Recent literature
Hu, L., Xu, J., Yin, M. X., Zhang, L., Lu, Y., Wu, W., Xue, Z., Ho, M. S., Gao, G., Zhao, Y. and Zhang, L. (2016). Ack promotes tissue growth via phosphorylation and suppression of the Hippo pathway component Expanded. Cell Discov 2: 15047. PubMed ID: 27462444
Non-receptor tyrosine kinase Activated cdc42 kinase (see Drosophila Ack) was reported to participate in several types of cancers in mammals. It is also believed to have an anti-apoptotic function in Drosophila. This study reports the identification of Drosophila Activated cdc42 kinase as a growth promoter and a novel Hippo signaling pathway regulator. Activated cdc42 kinase promotes tissue growth through modulating Yorkie activity. Furthermore, Activated cdc42 kinase interacts with Expanded and induces tyrosine phosphorylation of Expanded on multiple sites. A model is proposed that activated cdc42 kinase negatively regulates Expanded by changing its phosphorylation status to promote tissue growth. Moreover, ack genetically interacts with merlin and expanded. Thus, this study identifies Drosophila Activated cdc42 kinase as a Hippo pathway regulator.

The first expanded (ex) mutation to be identified (Stern and Bridges, 1926) causes wide wings. Waddington (1940) characterized the ex phenotype in greater detail and concluded that the wing defect is probably due to effects on cell division. An allelic series of ex mutations has been characterized that causes varying degrees of hyperplastic overgrowth of discs. expanded has now been cloned and characterized. It encodes a protein with a canonical N-terminal 4.1 homology domain, with three potential SH3-binding sites (Boedigheimer, 1993). The SH3-binding sites may serve as docking sites for SH3-containing proteins, such as Discs large, Drosophila Abl oncogene or Drosophila Src (Boedigheimer, 1993). The 4.1 homology domain mediates interactions with cell membrane-bound proteins. It is tempting to speculate that the localization of Ex to adherens junctions (Boedigheimer, 1997), where a large number of signaling and structural proteins are also localized, may allow it to exert effects on multiple processes involved in disc development, including effects on cell proliferation, cell fate determination, and tissue polarity (Blaumueller, 2000).

Merlin, another member of the Protein 4.1 superfamily, also plays a role in the the regulation of cell proliferation; proteins in this family were previously thought to function primarily to link transmembrane proteins to underlying cortical actin. Loss of Merlin function in Drosophila results in hyperplasia of the affected tissue without significant disruptions in differentiation. Similar phenotypes have been observed for mutations in expanded. Because of the phenotypic and structural similarities between Merlin and expanded, it was asked whether Merlin and Expanded function together to regulate cell proliferation. Recessive loss of function of either Merlin or expanded can dominantly enhance the phenotypes associated with mutations in the other gene. Consistent with this genetic interaction, Merlin and Expanded colocalize in Drosophila tissues and cells, and physically interact through a conserved N-terminal region (CNTR) of Expanded, characteristic of the Protein 4.1 family, and the C-terminal domain of Merlin. Loss of function of both Merlin and expanded in clones reveals that these proteins function to regulate differentiation in addition to proliferation in Drosophila. These results indicate that Merlin and Expanded function together to regulate proliferation and differentiation, and have implications for an understanding of the functions of other Protein 4.1 superfamily members (McCartney, 2000).

expanded plays a role in patterning of the eye, mainly at the level of planar polarity. Mutant exe1 clones exhibit penetrant phenotypes that are characterized by ommatidial chirality inversions, misrotations, and minor defects in photoreceptor differentiation. In order to determine whether these defects arise during early stages of development, an antibody against the Spalt protein, a marker for the R3 and R4 photoreceptor precursors (on which ommatidial rotation and chirality depend) was used to analyze rotation defects within mutant tissue in pupal imaginal discs. The random orientation of many of the R3/R4 photoreceptor precursor pairs within the clone at this early stage indicates that defects in planar polarity are likely to be a primary consequence of loss of ex function. ex does not affect the initiation of differentiation or the progression of the morphogenetic furrow (differentiation is evident in both imaginal disc and adult tissue), but ex does play a role in orchestrating the fine details of the ensuing cell fate specification and planar polarization events (Blaumueller, 2000).

Given the planar polarity phenotype of ex loss-of-function mutants, a test was made of the ability of ex alleles to genetically interact with gain-of-function genotypes generated by overexpression of components of the frizzled (fz) planar polarity pathway. The gain-of-function frizzled and disheveled (dsh) phenotypes (sev-fz, sev-dsh), and also sev-rhoAV14 and sev-racV12, have been successfully used in previous studies both to identify new components of the Fz/Dsh planar polarity pathway, and to genetically position known components with respect to others. This assay, dominant genetic modification of sev-fz, sev-dsh, sev-rhoAV14 and sev-racV12, was used to analyze in more detail the role of Ex in this process. These experiments did not reveal significant genetic interactions between ex and most of these genotypes. However, sev-dsh is dominantly enhanced by exe1, with an increase of ommatidia that are unscorable with respect to polarity because they lack one or more photoreceptors. This phenotypic modification could be achieved by several mechanisms, and is suggestive of complex cross-talk, either between multiple signaling pathways (dsh itself plays a role in multiple signaling pathways, including those represented by wingless, frizzled, and possibly Notch), or between signaling pathways and the mechanical processes required to carry out the instructions provided by the signaling pathways (Blaumueller, 2000).

Overexpression of Ex in the wing results in a tissue reduction phenotype, consistent with a growth inhibitory role (Boedigheimer, 1997). In order to test whether this role of Ex is conserved across disc types, adult eyes in which Ex was overexpressed using the UAS-GAL4 system were analyzed. Early expression was expected to affect cells anterior to the morphogenetic furrow (cells undergoing non-synchronized divisions prior to fate determination), whereas late expression would affect cells posterior to the morphogenetic furrow (cells dividing synchronously in the midst of differentiating cells). Externally visible adult phenotypes result from Ex expression under the control of a variety of early and late drivers including eyeless (ey), scabrous (sca), sevenless (sev), and glass multimer reporter (GMR). Early Ex overexpression throughout the developing eye disc under the ey driver (starting in embryonic development) results in a phenotype characterized by dramatic effects on the overall size and shape of the eye, as well as a mild roughening of the surface. Externally, eyes often bulge and are reduced in the ventral region. Retinal sections through these eyes reveal mild patterning defects that primarily affect the positioning of the equator, the border between the two chiral forms of ommatidia present in the dorsal and ventral halves of the eye. This defect is likely to be a secondary consequence of distortions of the eye disc (Blaumueller, 2000).

Overexpression of Ex under the remaining drivers, which are late-expressing, results in a distinct phenotype characterized by roughening and blistering externally, and patterning defects in sections. In the case of 32B (expression posterior to the morphogenetic furrow), external roughening of the eye reflects ommatidial misrotations and a loss of pigment cells (identifiable by yellow pigment granules). Similar effects are produced using the sca driver (expression in a small subset of cells posterior to the morphogenetic furrow). Overexpression of Ex under the control of either sev (transient expression in a subset of photoreceptor and all cone cell precursors), or GMR (expression in all cells within and posterior to the morphogenetic furrow throughout the remainder of development) results in severe effects on eye development. In both the latter cases, adult eyes are blistered and reduced in size. Sectioning through blistered regions reveal a mass of highly disorganized ommatidia, whereas the more mildly affected posterior regions of sev-ex eyes exhibit a phenotype similar to that seen in 32B-ex and sca-ex eyes. The most prominent features of this phenotype are a loss of pigment cells and defects in planar polarity. Occasionally, a photoreceptor cell is also missing within an ommatidium. The results are consistent with Ex overexpression causing an overall reduction in cell number. The small eye phenotype obtained by expressing Ex early under the ey driver can be explained by a reduction in cell number early in eye development. Likewise, the phenotypes generated by overexpression under the late drivers suggest that cell loss occurs, but at a point in time at which such effects impact tissue patterning as well as tissue size (Blaumueller, 2000).

An examination of eye imaginal discs in which Ex expression is driven early (ey-ex) reveals striking effects on disc size and shape. Discs were double-stained for markers of morphological differentiation (anti-Notch) and neuronal development (anti-ELAV). These stainings reveal that the eye imaginal discs are consistently small relative to the antennal portions of the same disc complex. Unlike wild-type discs, ey-ex discs are not flat. This effect could account for the misshapen adult eyes. In contrast to the dramatic effects on tissue size, those on cell fate specification and differentiation are mild, ommatidial preclusters looking fairly well-ordered. Together with the adult phenotype, these results indicate that the defects are caused mainly by early effects of Ex expression on cell number anterior to the furrow, prior to the time at which patterning events take place (Blaumueller, 2000).

An examination of discs in which Ex expression is driven posterior to the morphogenetic furrow reveals early defects in ommatidial polarity as well as effects on ommatidial precluster density and on cell number. The positions of specific subsets of photoreceptors were monitored using the svp-beta-galactosidase reporter (svp-beta-gal; expressed specifically in the R3/R4 photoreceptor precursors beginning just posterior to the furrow, and also in the R1/R6 precursors slightly further posterior in the disc), making it possible to analyze ommatidial polarity from the earliest point of its establishment. In wild-type discs, the R3/R4 and R1/R6 precursor pairs are aligned in precise diagonal rows. In the case of 32B-ex discs, this pattern is also fairly regular. However, several misrotated clusters are present, even in the anterior-most rows in which the R3/R4 pair can be identified. In the cases of sev-ex and GMR-ex discs, rotation defects are also prominent in anterior rows. Polarity could not be assessed in posterior regions of these discs because individual R3/R4 and R1/R6 pairs could not be distinguished in the dense mass of svp-beta-gal positive cells. In addition to its effects on planar polarity, Ex overexpression posterior to the furrow can lead to the loss of cells in which expression is driven. In sev-ex discs, one of the R3/R4 precursor pair is occasionally missing, consistent with the loss of photoreceptors in sections through adult tissue. Furthermore, the number of accessory cells that normally surround the photoreceptors appears to be greatly reduced. The adult phenotypic analysis suggests that both cone and pigment cells are affected: eye blistering is characteristically a consequence of defects in the cone cells that secrete the lens material, and pigment cells are clearly missing in retinal sections. Also, the svp-beta-gal stainings reveal that ommatidial preclusters are packed more densely than in wild-type discs. To test whether cone cells (in which sev also drives expression) are affected from early stages, sev-ex pupal discs were stained for the nuclear cone cell marker Cut. Whereas arrays of four Cut positive cells are present per precluster in wild-type discs, fewer are present in preclusters of sev-ex discs, and the ommatidial array is markedly disturbed. Hence, cone cell loss contributes to the phenotype seen in sev-ex flies, as does photoreceptor loss. These results are consistent with cell loss at, or after, the time of cell fate specification (Blaumueller, 2000).

The results of the experiments described above, coupled with previous loss-of-function and overexpression studies in the wing, suggest that ex regulates the growth of imaginal discs by inhibiting cell proliferation. The elimination of an inhibitor of proliferation in loss-of-function mutants could explain the overgrowth seen in these animals. Conversely, overexpression of such an inhibitor could account for the overall loss of tissue throughout the disc when expressed at early stages, and for the loss of late-born cells derived from the second mitotic wave when expressed at late stages. Moreover, phenotypes similar to those described for Ex overexpression posterior to the morphogenetic furrow have been obtained by inhibiting the second mitotic wave by overexpressing the cyclin-dependent kinase inhibitor p21 under the control of the GMR driver. The hypothesis that Ex overexpression negatively regulates proliferation was tested by staining for either the S-phase marker bromodeoxyuridine (BrdU), or the mitosis marker phospho-histone H3 in discs expressing Ex under the drivers discussed above. None exhibit a detectable change in the staining for either marker when compared to wild-type. Hence, it appears that Ex overexpression does not interfere with cell proliferation (Blaumueller, 2000).

This observation raises the question of whether Ex exerts its effects at the level of cell death. Discs overexpressing Ex were therefore tested for the incorporation of the cell death marker, Acridine Orange. Wild-type eye imaginal discs exhibit low levels of apoptosis both anterior and posterior to the morphogenetic furrow. In contrast, discs from both sev-ex and GMR-ex larvae exhibit dramatic increases in Acridine Orange incorporation. Moreover, the regions of incorporation overlap with the expression patterns of each driver. Whereas high levels of Acridine Orange incorporation are limited to a relatively narrow band just behind the morphogenetic furrow in sev-ex discs, in GMR-ex discs the staining pattern begins and ends further to the posterior. In the case of ey-ex, the incorporation of Acridine Orange in third instar larval discs is also increased, although this is more variable. Interestingly, in several ey-ex discs with high levels of Acridine Orange incorporation, the marker is regionalized to the ventral half, potentially accounting for the frequent loss of this region of the adult eye. These data demonstrate that Ex overexpression causes tissue reduction by inducing massive cell death. In order to determine whether this is also the case in other tissues, the effects of Ex overexpression on the wing and leg were tested. Misexpression of Ex using the optomotor blind (omb) driver, which is expressed throughout much of the region that gives rise to the wing pouch and also in a sector of the leg disc, results in discs incorporating high levels of Acridine Orange in a pattern overlapping that of omb-driven Ex overexpression. These flies have severely reduced wings with patterning defects including loss of vein material and disruption of the margin. Legs appears to be fragile and twisted. Similarly, overexpression using the MS1096 driver results in small, unpatterned wings and misshapen legs. The patterning defects generated by expression under both of these drivers probably reflect the extensive overlap of their expression domains with those of patterning elements for these tissues (Blaumueller, 2000).

The viral caspase inhibitor p35 is a known suppressor of apoptosis. Coexpression of this protein with Ex results in a dramatic rescue of the sev-ex phenotype. Externally, the eye is much smoother than in sev-ex and in sections most of the cell loss is rescued, although occasional photoreceptor and pigment cell loss is evident. Defects in planar polarity are still prominent. Taken together, the data presented here demonstrate that the tissue reduction phenotypes produced by Ex overexpression in all disc types are due to caspase-dependent, and therefore apoptotic, cell death (Blaumueller, 2000).

The most obvious consequence of loss of ex function in all imaginal discs is the deregulation of growth. Homozygous mutant ex clones have a significant growth advantage over their wild-type counterparts, and eye discs from larvae homozygous null for the gene are vastly overgrown. Hence, the apparent reduction of adult eye tissue originally observed in ex mutants may be a secondary consequence of a failure in disc eversion and morphogenesis of overgrown tissue. The overgrowth phenotypes in the eye are consistent with those seen in the wing (Boedigheimer, 1993; Boedigheimer, 1997). Like the loss-of-function phenotypes, those obtained with Ex over-expression are internally consistent. Regardless of disc type, Ex overexpression leads to tissue reduction that is a consequence of apoptosis in the region of overexpression, and its precise effects on size and patterning of the disc are dependent on its timing relative to decisions being made in the tissue (Blaumueller, 2000).

The finding that ex plays a common role in different disc types simplifies an understanding of the way in which this gene functions. However, the unexpected observation that the overexpression phenotype is a consequence of the induction of cell death, rather than of a repression of proliferation, complicates the issue. The most obvious model consistent with the overexpression data by itself would be that Ex normally acts as an inducer of apoptosis, and that loss of the protein leads to insufficient cell death in the disc, with consequent tissue overgrowth. However, apoptosis normally occurs at very low levels in larval imaginal discs and a lack of apoptosis on its own is not likely to cause the massive overgrowth that is evident in mutants as early as the third instar. This idea is supported by the fact that overexpression of the viral caspase inhibitor, p35 (which blocks apoptotic cell death in the eye), produces eyes that are only slightly larger than normal. If ex is required for controlling tissue growth, but is also capable of inducing apoptosis, how are these roles related to one another? One possibility is that Ex normally takes part only in regulating growth, but when expressed at high levels, interferes with growth and ultimately leads to apoptosis as a secondary effect. The fact that proliferation is not inhibited when Ex is overexpressed suggests that this is not the case. However, one cannot rule out the possibility of proliferation arrest at a late point within the cell cycle, which would not be detectable by BrdU or phospho-histone H3 labeling and might itself cause the cells to die. An alternative explanation is that Ex might play independent direct roles in both negatively regulating proliferation and positively regulating cell death. Its role as a negative regulator of growth might more easily be unmasked by loss-of-function mutations than by overexpression; Ex might be necessary for preventing overgrowth, but insufficient for inhibiting normal proliferation when overexpressed. Conversely, a role as a positive regulator of cell death might easily be detected when it is enhanced by overexpression, but not in the loss-of-function situation, where it could be masked by the overgrowth phenotype (Blaumueller, 2000).

Ex appears to play a direct role in certain aspects of tissue patterning. Both loss-of-function and overexpression mutants exhibit some patterning defects that cannot be explained as secondary effects. For example, in the case of Ex overexpression in sev-ex, rotation defects are present early in disc development, prior to the death of the cells in which it is expressed. Hence, this defect in planar polarity must be a direct consequence of Ex overexpression. The fact that the planar polarity defects exhibited by ex loss-of-function and overexpression mutants resemble one another, as well as being similar to those of previously characterized planar polarity genes, further supports the notion that Ex plays a direct role in this process. However, due to the lack of informative genetic interactions with components of the Fz pathway, the role of ex in planar polarity establishment remains unclear. It is possible that the observed phenotypic interaction between ex and sev-dsh reflects cross-talk between multiple signaling pathways (dsh itself has been implicated in several signaling pathways, or between signaling pathways and the mechanical processes involved in generating ommatidial polarity. However, regardless of whether ex plays a role in promoting cell signaling or in initiating the morphogenetic movements required for establishing planar polarity, its overall importance in organizing planar polarity is an exciting finding (Blaumueller, 2000). The possibility that Ex plays a role in the mechanics of planar polarity is particularly intriguing, since nothing is yet known about how the signaling events controlling planar polarity are translated into the physical movements that take place to establish this phenomenon in the eye. Hence, ex could provide a unique entry point en route to finding a solution for this problem. With regard to the potential connection of signaling and mechanical events by Ex, it is noted that the loss of ex does not lead to an obvious perturbation of the cellular architecture (as ascertained by staining of the eye disc for the junctional markers Notch (adherens junctions) and Coracle (septate junctions), suggesting that ex is not a global regulator of junctional structure. However, when overexpressed, Ex is mislocalized: in addition to being concentrated in apical regions of the membrane, the protein is also diffusely distributed throughout the cytoplasm. Hence, defects in differentiation, a process that relies heavily on cell-cell contacts, may arise as a consequence of Ex sequestering important signaling or structural components away from the appropriate junction, or perhaps even by allowing these molecules to act at inappropriate times and places. Certainly, this mislocalization could also account for other effects of Ex overexpression, for example, the induction of apoptosis (Blaumueller, 2000).


cDNA clone length - 6696

Bases in 5' UTR - 891

Exons - 6

Bases in 3' UTR - 905


Amino Acids - 1429

Structural Domains

The Drosophila expanded gene encodes a product that shares homology with the protein 4.1 family of proteins (Boedigheimer, 1993). Ex protein contains potential SH3-binding sites. Comparisons of the predicted Ex protein sequence to protein databases and translated DNA databases reveals no striking similarities to known proteins. A lack of hydrophobic segments suggests that the Ex protein is probably not secreted or membrane spanning. Overall the predicted protein is basic, with three positively charged amino acid segments (isoelectric point greater than 9.0) separated by neutral or slightly acidic regions. The C-terminal half of the protein contains many short homopolymeric runs of proline, serine, alanine, histidine and glutamine. Three consensus binding sites are present for Src homology 3 (SH3) domains, defined as PXXPPPXXP. SH3-binding sites have also been found in 3BP1, a mammalian protein that also has homology to GAP-rho outside the SH3-binding site, and in formins, which are products of the limb deformity gene in mice, and in the muscarinic acetylcholine receptor genes. Homopolymeric runs of glutamine, proline and histidine have been associated with protein-protein interactions in transcription factors. Ex also contains QA and LX repeats of unknown significance. Searches of Prosite reveal 51 potential serine-threonine phosphorylation sites and 1 potential tyrosine phosphorylation site (Boedigheimer, 1993).

expanded: | Regulation | Developmental Biology | Effects of Mutation | References

date revised: 25 June 2000

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.